Overview
Xu Lin is an AI engineer and researcher whose public work mainly sits at the intersection of multimodal models, computer vision, evaluation, and applied engineering. Basic Profile - Name: Xu Lin - Current base: Shenzhen - Current organization: Tencent Youtu Lab - Main areas visible in public work: multimodal models, evaluation, computer vision, agent-related engineering - Email: islinxu@163.com - GitHub: github.com/isLinXu What This Site Represents This site is not meant to claim every possible capability. It mainly collects the work that is easiest to verify publicly: - repositories and open-source contribution - paper and benchmark related output - technical notes and curation work - a web resume version based on recent materials Current Direction Recent public repositories and profile material point to several recurring directions: - multimodal large-model related work - evaluation and benchmark-oriented thinking - practical computer vision engineering - agent-oriented tools and workflows Working Style - comfortable moving between research reading, implementation, and tooling - values evaluation and reproducibility over vague claims - tends to turn ongoing learning into public repositories, notes, or curation projects Representative Public Work - paper-list - YOLOv8 Efficient - crablet - claude-code-book - LlamaFactory - MME related benchmark work Public Links - GitHub: github.com/isLinXu - LinkedIn: linkedin.com/in/xu-lin-3b78a5251 - Blog: www.cnblogs.com/isLinXu - Hugging Face: huggingface.co/gatilin Note The purpose of this page is to give a concise and verifiable summary. More detailed project and achievement information is split across the Projects , Highlights , and Publications pages.